Estimating prevalence of injecting drug users and associated heroin-related death rates in England by using regional data and incorporating prior information

Ruth King, Sheila M Bird, Antony M. Overstall, Gordon Hay, Sharon Hutchinson

Research output: Contribution to journalArticle

17 Citations (Scopus)

Abstract

Injecting drug users (IDUs) have a direct social and economical impact, yet can
typically be regarded as a hidden population within a community. We estimate the size of the IDU population across the nine different Government Office Regions of England in 2005/6 using capture-recapture methods with age (ranging from 15-64) and gender as covariate in-formation. We consider a Bayesian model-averaging approach using log-linear models, where we are able to include explicit prior information within the analysis in relation to the total population size (elicited from the number of drug-related deaths and injectors’ drug-related death rates) and the male to female ratio of IDUs. Estimating the data at the regional level allows for regional heterogeneity and was aggregated to obtain an estimate at the England level with posterior mean of 194600 and 95% credible interval (180350, 208800), estimated to nearest 50. The results show significant regional variability in the estimated prevalence of current IDUs (with posterior means ranging from 3 to 9 per 1000 of population aged 15-64) and injecting
drug-related death rates across the gender age cross-classifications.
LanguageEnglish
Pages209-236
Number of pages28
JournalJournal of the Royal Statistical Society: Series A
Volume177
Issue number1
Early online date23 Apr 2013
DOIs
Publication statusPublished - Jan 2014

Fingerprint

death rate
Prior Information
Drugs
drug
Posterior Mean
government office
Bayesian Model Averaging
Credible Interval
Capture-recapture
Injector
Log-linear Models
gender
linear model
Population Size
Estimate
Prior information
Heroin
Death rate
England
Injecting drug users

Keywords

  • drug-related deaths
  • log-linear models
  • population size
  • injecting drug users
  • model-averaging
  • prior information

Cite this

@article{2cb520efe3b147dab5f24c0f47c6d7c4,
title = "Estimating prevalence of injecting drug users and associated heroin-related death rates in England by using regional data and incorporating prior information",
abstract = "Injecting drug users (IDUs) have a direct social and economical impact, yet cantypically be regarded as a hidden population within a community. We estimate the size of the IDU population across the nine different Government Office Regions of England in 2005/6 using capture-recapture methods with age (ranging from 15-64) and gender as covariate in-formation. We consider a Bayesian model-averaging approach using log-linear models, where we are able to include explicit prior information within the analysis in relation to the total population size (elicited from the number of drug-related deaths and injectors’ drug-related death rates) and the male to female ratio of IDUs. Estimating the data at the regional level allows for regional heterogeneity and was aggregated to obtain an estimate at the England level with posterior mean of 194600 and 95{\%} credible interval (180350, 208800), estimated to nearest 50. The results show significant regional variability in the estimated prevalence of current IDUs (with posterior means ranging from 3 to 9 per 1000 of population aged 15-64) and injectingdrug-related death rates across the gender age cross-classifications.",
keywords = "drug-related deaths, log-linear models, population size, injecting drug users, model-averaging, prior information",
author = "Ruth King and Bird, {Sheila M} and Overstall, {Antony M.} and Gordon Hay and Sharon Hutchinson",
note = "details updated",
year = "2014",
month = "1",
doi = "10.1111/rssa.12011",
language = "English",
volume = "177",
pages = "209--236",
journal = "Journal of the Royal Statistical Society: Series A",
issn = "0964-1998",
number = "1",

}

Estimating prevalence of injecting drug users and associated heroin-related death rates in England by using regional data and incorporating prior information. / King, Ruth; Bird, Sheila M; Overstall, Antony M.; Hay, Gordon; Hutchinson, Sharon.

In: Journal of the Royal Statistical Society: Series A , Vol. 177, No. 1, 01.2014, p. 209-236.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Estimating prevalence of injecting drug users and associated heroin-related death rates in England by using regional data and incorporating prior information

AU - King, Ruth

AU - Bird, Sheila M

AU - Overstall, Antony M.

AU - Hay, Gordon

AU - Hutchinson, Sharon

N1 - details updated

PY - 2014/1

Y1 - 2014/1

N2 - Injecting drug users (IDUs) have a direct social and economical impact, yet cantypically be regarded as a hidden population within a community. We estimate the size of the IDU population across the nine different Government Office Regions of England in 2005/6 using capture-recapture methods with age (ranging from 15-64) and gender as covariate in-formation. We consider a Bayesian model-averaging approach using log-linear models, where we are able to include explicit prior information within the analysis in relation to the total population size (elicited from the number of drug-related deaths and injectors’ drug-related death rates) and the male to female ratio of IDUs. Estimating the data at the regional level allows for regional heterogeneity and was aggregated to obtain an estimate at the England level with posterior mean of 194600 and 95% credible interval (180350, 208800), estimated to nearest 50. The results show significant regional variability in the estimated prevalence of current IDUs (with posterior means ranging from 3 to 9 per 1000 of population aged 15-64) and injectingdrug-related death rates across the gender age cross-classifications.

AB - Injecting drug users (IDUs) have a direct social and economical impact, yet cantypically be regarded as a hidden population within a community. We estimate the size of the IDU population across the nine different Government Office Regions of England in 2005/6 using capture-recapture methods with age (ranging from 15-64) and gender as covariate in-formation. We consider a Bayesian model-averaging approach using log-linear models, where we are able to include explicit prior information within the analysis in relation to the total population size (elicited from the number of drug-related deaths and injectors’ drug-related death rates) and the male to female ratio of IDUs. Estimating the data at the regional level allows for regional heterogeneity and was aggregated to obtain an estimate at the England level with posterior mean of 194600 and 95% credible interval (180350, 208800), estimated to nearest 50. The results show significant regional variability in the estimated prevalence of current IDUs (with posterior means ranging from 3 to 9 per 1000 of population aged 15-64) and injectingdrug-related death rates across the gender age cross-classifications.

KW - drug-related deaths

KW - log-linear models

KW - population size

KW - injecting drug users

KW - model-averaging

KW - prior information

UR - http://www.scopus.com/inward/record.url?scp=84876667446&partnerID=8YFLogxK

UR - http://lemur.mcs.st-and.ac.uk/mypapers/kinbhh10.pdf

U2 - 10.1111/rssa.12011

DO - 10.1111/rssa.12011

M3 - Article

VL - 177

SP - 209

EP - 236

JO - Journal of the Royal Statistical Society: Series A

T2 - Journal of the Royal Statistical Society: Series A

JF - Journal of the Royal Statistical Society: Series A

SN - 0964-1998

IS - 1

ER -